"Whenever I go into a restaurant, I order both a chicken and an egg to see which comes first"

Friday, December 21, 2012

Should We Pay More For Our Risky Behavior?

On the surface, paying more to insure risky behavior seems like a good idea.  Why should I pay a high health insurance premium which has been driven up by smokers and fat people? Or why should I have sky-high car insurance rates because the coverage pool includes joy-riding teenage boys?

Yet, there is an inherent danger in this trend.  Once insurance companies catch on to consumer’s enthusiastic response to variable pricing, they will happily find ways to expunge all high-risk individuals from their population pool.  Thanks to big data, their job has been made easier, and there are limitless possibilities to purge the ranks.  Big data can now facilitate the correlation between many psycho-social factors that once were beyond scientific reach.  Small data have already produced obvious correlations.  Everyone knows that age is correlated with disease and traffic accidents.  If you are past 70 you will soon catch something nasty and die; and if you are over 80 and still driving, the chances of getting into a bad wreck are very high.

Other correlations which were not possible before, are now common thanks to big data.   Psychological profiling, made possible by collecting hundreds of millions of Facebook posts and Twitter feeds, sifting them through the tight sieves of ‘sensitivity’ algorithms to pick up clues about erratic behavior allows insurance company to correlate that reported behavior with actual traffic accidents or tickets.  While individual information is still proprietary (although not for long), researchers can match expressed activities and the emotional affect attached to them with groups of individuals identified by age, place of residence, occupation, etc.

Tracking emails that confide health information can be useful in terms of adding a predictive element to biostatistics. It is easy to measure and quantify illness once it happens; but predicting it is more difficult and more rewarding.  If an insurance company has correlated disease-related symptoms discussed on the Internet with eventual high rates of illness, it can choose to eliminate classes of individuals because of their higher risk.  As above, the law currently does not permit individual profiling, but an insurance company can aggregate data so that it accurately describes any collection of individuals.

This privacy firewall, however, will not last long.  Just as Amazon.com can precision market thanks to computer cookies; and Gmail can sell information on your retail preferences gleaned from emails to marketers, soon insurance companies will be able to buy those data and determine if you personally are a bad risk.  Insurance companies would be delighted to know if you have searched cancer sites over the past weeks, and can track to see when you were worried about diagnosis and when you switched to treatment.  Again, the insurance companies can add a predictive measure to their existing data.

This is the nature of the private insurance industry.  In order to make a profit, they have to take in far more money in premiums than the pay out in benefits.  If they can create a client list of insured who are all low-risk, then they will increase their profits.

A national health care system – or a no-fault car insurance system – would eliminate actuarial science.  That is, it would pool all Americans or all drivers within a single, large pool.  The healthy will always outnumber the sick; the good drivers will always outnumber the bad.  As it stands, the system is engineered to discriminate against risk.

The conservative objections to nationalized health care have been made well-known.  If an individual wants to smoke and add to his health risk, let him do so, but let him pay for it, and let private insurance companies broker the deal.  Why should all Americans – especially responsible, risk-averse, God-fearing citizens – have to assume the burden of life’s profligates?  Why should they contribute one cent to the wayward, dissolute behavior of drug addicts, sexual perverts, undisciplined cake-eaters, and sky-divers?  It is a matter of principle.

This attitude, however, has negative consequences.  While anything that encourages upright, proper living should be encouraged – and in this case the marketplace can be very effective through variable pricing in doing so – many people are at higher risk for no behavioral fault of their own.  Current scientific theory states that certain dysfunctional behavior, like alcohol or drug addition, is genetically linked.  In other words, our ability to regulate our behavior is limited.  Theories on obesity suggest that there may be a genetic link there as well.  Should we as a society condemn these unfortunate people to the margins of a healthy society?

The recent accuracy and popularity of DNA testing has raised the ante.  Insurance companies now have scientifically reliable data to allow them to refine their predictive algorithms.  If a woman has a particular genetic variant in her DNA her chances of getting breast cancer or ovarian cancer can be calculated with great accuracy.  It is no longer a question of whether you might get breast cancer because your aunt died of it; it is bald statement that you have a 78 percent risk of getting it. Once insurance companies get a hold of that information, rates will go up.  Way up.

Although under Obamacare no coverage will be denied because of pre-existing conditions, there is nothing to stop insurance companies from insisting on genetic testing and adjusting premium levels according to the results.

More importantly, this ‘throw the sick to the side of the road’ mentality has ethical and moral consequences.  Is there any reason why a woman who has inherited Great-great Aunt Mabel’s twisted cancer gene should be penalized for that 100 year-old mutation? Even without considering genetic pre-disposition, shouldn’t we be more generous and tolerant as a nation.  So what if Uncle Harry was a boorish, drunken lout who died of cirrhosis of the liver at 55.  His wife, Agnes, was a saint. It all balances out in the end.

An editorial in The Telegraph (12.21.12) informs us that variable pricing will be the law in the UK from now on, and the staff thinks it is a good idea:

Insurance should be based on risk, not human rights. In any case, men and women drivers after the age of 40 tend to be treated equally already. The reason the premiums vary so much for lower age groups is because young men have a far greater proportion of serious accidents.

Risk analysis, prediction, and variable pricing are here to stay.  We have become too individualistic a society to even consider the communitarian aspects of public policy.  We have also become too moralistic and judgmental. This is not a good thing.

No comments:

Post a Comment